ACSL4-driven ferroptosis susceptibility as a targetable vulnerability in monocytic acute myeloid leukemia

Ferroptosis, an iron-dependent form of regulated cell death driven by lethal lipid peroxidation, has emerged as a targetable vulnerability in cancer. ACSL4 is the rate-limiting enzyme that dictates ferroptosis sensitivity by channeling polyunsaturated fatty acids into membrane phospholipids. In acute myeloid leukemia (AML), monocytic subtypes resist BCL2 inhibition with venetoclax, yet their metabolic dependencies remain poorly defined. Here, we integrated PRISM drug-sensitivity data (6,790 compounds) and DepMap CRISPR dependencies (18,435 genes) across 15 adult AML cell lines to map drug-gene co-dependencies. A multi-cohort validation strategy - filtering 50 discovery candidates (29 testable in BeatAML) through ex vivo drug-response data in 476 primary AML specimens and clinical outcomes in 140 adults with de novo AML (TCGA-LAML) - converged on a single axis linking SRC family kinase inhibition to ACSL4 expression. ACSL4-high blasts showed enhanced dasatinib sensitivity (r = -0.25, P = 4.3 x 10^-8). A composite SRC/ACSL4 signature stratified overall survival (HR 1.27; 95% CI 1.10-1.47; P = 0.0014), remaining significant after age adjustment. Single-cell atlas projection localized this signature to the monocytic compartment. The SRC/ACSL4-high state displayed a ferroptosis gene expression profile characterized by co-upregulation of ACSL4, HMOX1, and LPCAT3 with failure to upregulate the principal ferroptosis defense axis GPX4/SLC7A11. Conversely, ACSL4-high blasts showed significant ex vivo resistance to venetoclax (r = 0.36, P = 2.5 x 10^-12), linking the ferroptosis-primed monocytic state to BCL2 inhibitor failure. These findings nominate ACSL4-driven ferroptosis susceptibility as a lineage-specific vulnerability rendering monocytic AML selectively sensitive to SRC-directed therapy while resistant to BCL2 inhibition.